I'll try to respond to this, since Michael is probably quite busy with
the just-concluded meeting in Toronto, and since I'm interested in Old
Hungarian as well.
Please draw Michael Everson's attention to this message,
it may be helpful for clearing up some issues about proposed
Old Hungarian Runes, of which he seems to be in charge.
He wrote the proposal documents, N1686 and N1758, if that makes him in
charge of them.
His draft about the topic
http://www.dkuug.dk/jtc1/sc2/wg2/docs/n1758.pdf
says, that the spelling of character names for the proposed script
still needs to be determined, and two alternatives are presented, one
prefixes the names of most consonants with a letter e, the other
reflects the letter names of the Latin-based Modern Hungarian
alphabet, usually with a following vowel.
This insecurity
Probably more like uncertainty, but anyway:
seems to arise from the rather confusing English
explanations about letter names on webpages, like
http://fang.fa.gau.hu/~heves/runic.html
But if one is able to read the Hungarian sections of the same website,
like
http://fang.fa.gau.hu/~heves/szabalyok.html
http://fang.fa.gau.hu/~heves/abc/abc.html
it becomes evident, that Hungarian Runes, besides being alphabetic
signs in the first place, may also be used as syllabics in order to
save space, in which case the forms with a _preceding_ vowel are used.
This also accounts for having the parallel letters
OLD HUNGARIAN LETTER EK
and
OLD HUNGARIAN LETTER AK
and also
OLD HUNGARIAN LETTER ES
and
OLD HUNGARIAN LETTER AS
The good thing is that character names are not prescriptive. What a
character ends up being called does not influence or restrict its
potential usage.
(Actually, by some mistake the latter seems to be absent from the
present proposal. I hope these data are a convincing argument, that
it is actually necessary to include it. A .gif of the letter form
(as.gif) is accessible on the page
http://fang.fa.gau.hu/~heves/abc/abc.html )
AS was listed as a ligature in the earlier proposal, N1686. The later
document, N1758, proposed removing all the precomposed ligatures and
encoding them with sequences of letters instead. This led to an
interesting discussion in which Michael proposed a Zero-Width Ligator
character, a role that was eventually assigned to U+200D ZERO WIDTH
JOINER (though some apparently still feel ligation is a font and
application issue rather than a character encoding issue).
Bear in mind that N1686 and N1758 are more than six years old. There
have been no new proposals or other documents on rovsrs since 1998,
probably because additional research would require additional money that
is not currently available. Contributions to the Script Encoding
Initiative http://www.unicode.org/sei/ might help alleviate this
problem.
On the other hand, the names following the Modern Hungarian pattern
are only a kind of a shorthand, and have no claim to be part of the
Hungarian Runic writing system. (At least that's what the sites quoted
suggest.)
I can send Mr. Everson the relevant sections with an English
translation from these webpages if he needs them, but it would be to
long to send it all to the list.
They might well be helpful for determining the best names.
One more remark about letter names:
The long-vowel counterparts of the Umlaut letters
OLD HUNGARIAN LETTER OE and OLD HUNGARIAN LETTER UE
are distinguished in the proposal by doubling the second letter of the
digraph (OEE and UEE). It would blend better with Hungarian spelling
conventions and would probably be more straitforward to decipher, if
this was done by doubling the first element (OLD HUNGARIAN LETTER OOE
and OLD HUNGARIAN LETTER UUE), like in standard Hungarian compound
letters (e.g. cs (pron. ch, like in cheese), when pronounced
long, would be written ccs). This latter problem is of a different
nature though, and of relatively minor importance.
Again, this might be useful for determining the best names for these
letters.
Another remark about directionality:
Mr. Everson writes, that the impression I get is that the scholars
are used to LTR because it has been practical to implement on
computers and is less troublesome to read for people used to reading
the Latin script - but it seems that those particular needs should be
met with the directional overrides. This wording is not completely
accurate, because glyph shapes should also be mirrorred along the
vertical axis, depending on the writing direction.
Use of directional overrides often does imply mirroring of glyph shapes.
This is true for Old Italic, for example.
Note that Michael wrote in N1686, The glyphs presented in the tables
below are RTL glyphs; when LTR directionality is used (via UCS
directional modifiers), the glyphs must be reversed. You really do
want to read N1686 in conjunction with N1758:
http://std.dkuug.dk/JTC1/SC2/WG2/docs/n1686/n1686.htm
-Doug Ewell
Fullerton, California
